Slashdot Mirror
Is String Theory Really a Scientific Theory?
vk38 writes, "The New Yorker is running a story on whether String Theory is really a scientific theory or just an abstract exercise in math designed to churn out papers and Ph.Ds for the established academics. The article reviews two current books, by Lee Smolin and Peter Woit, laying out the case against string theory."
From the article: "Dozens of string-theory conferences have been held, hundreds of new Ph.D.s have been minted, and thousands of papers have been written. Yet... not a single new testable prediction has been made, not a single theoretical puzzle has been solved. In fact, there is no theory so far — just a set of hunches and calculations suggesting that a theory might exist. And, even if it does, this theory will come in such a bewildering number of versions that it will be of no practical use: a Theory of Nothing... String theory has always had a few vocal skeptics... Sheldon Glashow, who won a Nobel Prize for making one of the last great advances in physics before the beginning of the string-theory era, has likened string theory to a 'new version of medieval theology,' and campaigned to keep string theorists out of his own department at Harvard. (He failed.)"
String theory is a scientific theory that has neither been proved nor disproved to my knowledge.
I could speculate all day on whether or not it is fact but from what I've read, I will make a few statements. It seems that string theory was invented to satisfy some things we could not explain. This doesn't mean it's wrong or right although some people will contend that it is most probably wrong.
As the summary points out, few (if any) of String Theory's propositions can be tested or even observed. So it is simply an unknown right now. We cannot measure the proposed strings so how can we prove if they exist or they don't? We simply can't yet.
A good analogy would be Bohr's early assumptions about the atom. They were wrong but they were a step in the right direction. In hindsight, we see this now but we don't know what the future holds for String Theory. I'm just glad there are people out there thinking outside the box.
Do not fret, however, as scientists have been very resourceful at proving/disproving theories. I submit, for example, the exercise of determining the diameter of the building blocks of matter. Scientists had the idea to fill up one cubic milliletre of oil and dump it on top of a trough of water with a roller across the top. As the oil spread out, they moved the roller further down the trough. Once they started to see non-reflective parts of the water, they moved it back until they agreed the oil was completely spread out to the best of their abilities. Using this area, they determined how thick a molecule of oil could be without precision tools!
Similar ingenious tests have been devised to easily find the diameter of the earth at sunset on a beach with a yard stick or ruler.
So even though we may never be able to measure these strings, there are still some options left to explore to record properties that may prove/disprove their existence. We're merely in the very early stages of the scientific process.
Let us be excited about String Theory, even if it is wrong it sure is interesting. Nothing's wrong with a scientist who dreams, is there?
My work here is dung.
I am a geek, but I have seriously problems with math ability. I have a Bachelor of Arts degree. However, I like to keep up on math and science news as much as possible, inasmuch as I can understand it.
IIRC, string theory *does* make predictions, but the amount of energy required to run an experiement would be literally almost astronomical, so we have no practical way of testing it. I think according to concensus on what the 'scientific method' is, that makes it a hypothesis -- an educated guess, based on evidence. After it has sucessfully passed a few rounds of experiment, then we can say that it is a theory.
So, bottom line, it is scientific, as much as any other hypothesis. However, it's not a theory.
Computers are useless. They can only give you answers.
-- Pablo Picasso
String theory sounds weak. Let's upgrade the name so it sounds like it has to be true. Henceforth it will be referred to as String Fact.
I'll even throw a bone to an entrepreneural slashdotter out there. STRINGFACT.COM is not registered yet. It is yours for the taking.
Ninjas don't carry tic tacs
So which is it: the best of times or the worst of times?
According to Schrodinger, both.
Finkployd
Short Answer: No.
Long Answer: Yes.
Longer Answer: Both of the above, but each in a separate Universe.
An Indian-American Hindu committed to non-violent thought/speech/action alarmed by the global explosion of radical Islam
There's a difference between what's not practically testable and what's truly unfalsifiable. As long as it's conceptually possible to come up with a falsifying experiment, even if it's wildly impractical, it's still a scientific theory. We may yet come up with ways to test the theory. Sometimes that's because somebody comes up with a clever new test, an ingenious new reformulation of the theory, receives unexpected results from an exsting accelerator, or builds a new particle accelerator.
What's happening here is that people are complaining that the scientific establishment has made it difficult to work in alternatives to string theory. But just because you can't get a job to disprove a theory doesn't make it unfalsifiable. There needs to be healthy debate in the scientific community about which theories to work on. Shutting valid theories down is not healthy for science, but neither are accusations that conflate "impractical" with "impossible".
It's philosophy.
String theory is at the moment, philosophy. As soon as someone comes up with a way of testing it, it will become science.
Deleted
Let the physicists, who are the only people who can truly understand this, sort it out. They likely don't need the academic process becoming any more politicized than it already it. If it's a blind alley, they'll find that out in due time. While it's regrettable that it's taking as long as it is to reach a conclusion on the issue, come on - it ain't exactly flippin' burgers, and we're not exactly hung up waiting for the result. Let the scientists work.
there is no need to sign your posts. this isn't usenet. your username is right there above your post. stop it.
You can read about him on Wikipedia, if you like.
Ben Hocking
Need a professional organizer?
String theory is at times one of the biggest con jobs in Physics, and at other times some of the most interesting speculation. It's also the 'Theory that will NOT die!' reminding me of so many late night C rate thrillers.
Why? Because everytime string theory gets disproven, they come out with a new theory and call it 'String Theory'. String Theory from the 70's really doesn't resemble current string theory much other than the name. It's strange that this is so, but there are a lot more politics involved than there is science at times. And the author is right, there are lots of articles being written, but not much going on that can be said to prove the theory, and little in the way of predictions (cause those could be tested). And so far, everytime someone does stand up and make predictions, it quickly gets disproven by actual tests. Which may be why no one is predicting much using it anymore.
At this point actually String Theory may very well be the most 'disproven' theory in physics. But that doesn't seem to stop people from trying. It will be curious to see what science has to say about all of this 50 years from now. To be honest I think many of us have gotten too close to the subject to be objective about it, and I think that is not helping the issue on either side.
Having a "rolled up" dimension doesn't require an extra dimension, because they're not _actually_ rolled up. The metric used to describe them is just easy to picture that way. Just like curved 4-dimensional space time doesn't need a 5th dimension to be curved into. I tried looking for a good web-site that explains this, but didn't find one in the time I'm willing to spend looking for one. I'm sure someone else knows of one, though.
Ben Hocking
Need a professional organizer?
String theory might not have earned the rights to be called a theory yet, but as with Bohr's model of the atom, perhaps we could agree that it has earned the right to be called a model.
Ben Hocking
Need a professional organizer?
Seriously, the case may be that the reality of the Universe is so complicated that String Theory will take a long time to come to fruition. Also, it may be that there are no testable predictions because of our limited perspective.. eg 3 dimensions.. limited energy resources.. Fundamentally limited abilities to measure..
Or maybe the Universe is just a big knot of strings, and no human can untie it.
I am fud.
A Good Troll is better than a Bad Human.
Not only is there no way to test string theory at the moment, string theorists cannot even concieve of a way to test it in the future. As others have stated, the only tests people have thought of involve energy levels similar to that of the entire universe, to effect a change on an atomic scale.
So we have a "theory" that doesn't make any predictions, and cannot be tested. In that way it is very similar to "Intelligent Design" which also doesn't make any predictions, and cannot be tested. If ID isn't science, why is string theory?
While some of the math might be right, the same theory applies to friggin role playing games, too. So, are those real just because their math ads up?
Where are the good string theory experiments? Nowhere.
I scream. You scream. I assume that means we're both acquainted with the problem. We proceed.
http://en.wikipedia.org/wiki/Occam's_razor
:)
As far as my popular understanding of the domain goes, it goes like this. Before there was quantum phyiscs. Scientists thought lets smash these atoms see if there is anything inside them. So to the dismay of theirs they have been rewarded with millions of particle types quarks, muons etc etc. that they are trying to categorized catalogue, derive properties of. Some of them didn't like the idea that millions of disjoint test results as material for explaining universe's compositions. With advances in field of mathematics and nod from those early einstein papers they moved on trying produce the theory of everything. Sort of like beautiful theory of relativity. Though relativity has been easy to test and formulas are often recognized by some 6th 7th grade students (E=mc^@), string theory is quite a bit more complicated then that. As it stands of nearly infinite data result domain of quantum physics.
As the string theory suggests that protons neutrons and electrons are singlewaveforms of certain frequency. And smashed atoms and half-waveforms and for some reason decay rapidly.
I suppose it is an excersize in occam's razor placed into the future when theory can be verified.
Why scientists are folling said theory, is in their wet deams they think of Unified field theory, which string theory may well support.
Just like way back as someone mentioned here Bohr's suppositions were incorrect in many ways, but generally incorrect. Perhaps string theory will inspire a new one in the future, that will make more sense.
But for now I would think it should be renamed a hypothesis, away from shameless marketing of non existant product!
2c.
Mr. Glashow can be a genious in the field of Physics, but I doubt he's also so much of a genious in the fields of History, Philosophy and (yes) Theology to be able to make such an absurd statement. No matter how much he dislikes religion and related subjects, there's a difference between stating a personal taste and talking meaningfully about something you don't know about.
Conservatism: (n.) love of the existing evils. Liberalism: (n.) desire to substitute new evils for the existing ones.
As noted at http://web.lemoyne.edu/~giunta/whewell.html, an excerpt of a text by William Whewell from Philosophy of the Inductive Sciences vol. 1, 1840, pp. 406-7 [from Maurice Crosland, ed., The Science of Matter: a Historical Survey (Harmondsworth, UK: Penguin, 1971)]
He then went on to say that it could never be proven and would remain a work of philosophy and a tool for efficient calculation only.
I don't think it's smearing science, indeed in some ways it's keeping intact science's integrity in the face of "ID" and other anti-science movements.
String theory appears, for the most part, to be a very smart, very compelling, system that could be used to explain how the universe works. As such, it's easy to get lost in the excitement and forget that the current evidence for it is, well, not what it could be.
The author is saying "We should hold off and be careful about how we portray this, especially in relation to other scientific principles like relativity. It clearly isn't in the same class." That's absolutely right to do, and it helps prop up the scientific method if there's this kind of auditing going on all the time. It's no more defaming science than it is for someone to come up with a new "test" for relativity, who then does that.
You are not alone. This is not normal. None of this is normal.
it's true, it really is FUD.
String theory hasn't been replaced by newer versions, it's been updated with small modifications like "what if the basic premise is the same, but instead of a 1D string vibrating in 4 dimensions (x,y,z and t) it's vibrating in 11 dimensions, where the other dimensions are curled up within the planck length?"
There are reasons why string theory has failed to come up with any NEW predictions. For one thing, it's being constantly tweaked so that it is consistent with EXISTING experimentation. After all, why would you build a theory that you hope will become a GUT if it's not consistent with other proven theories?
The other thing is that this is a theory... the fact that it (mathematically) treats particles as being a 1D string vibrating in n-dimensions doesn't actually mean that if you could see items smaller than the planck length, that you would actually see a vibrating string!! It's a mathematical representation... the math doesn't have to represent exactly what's happening as long as it can be used to describe what is happening.
After all, modern chemistry is incredibly useful for predicting how atoms interact with eachother to form compounds... even though it's based off the idea that electrons orbit a nucleus like a tiny little planet orbitting a sun... that is precisely NOT what an electron does, but who cares, the math allows you to make determinations. It's the same with string theory.
I do not think that string theory is a con job. I do, however, think that attempting to come up with a GUT is a MUCH MUCH larger task than simply trying to explain, say, quantum behaviour, like tunneling.
They're starting with a very simple, and very elegant premise (that all particles are periodic vibrations with different frequencies corresponding to different particles) and then building from there. Hell... start with that and just try and figure out how to represent the periodic table... that alone would be mind-boggling. Now start trying to figure out what particle interaction would look like... then build up from there. The trick, is that it's possible to describe nearly everything using this theory... but it hasn't happened yet. Just because it's possible doesn't mean it'll be easy.
This, of course, probably means that it's the wrong way of going about it... but that doesn't make it a waste of time... the hardest part, I think, will be in having enough patience to see what the theory can produce outside of existing theories... unfortunately it has to be harmonized with existing theory ;-)
Oh god, that woman is John Romero!
Insolubility of the quintic, Godels Incompleteness Theorem ..... Sure it's possible to prove something can't be done. Maybe Scott Adams isn't the best source to base your view of logic.
Of course String Theory makes testable predictions. Just like General Relativity and Quantum Mechanics make testable predictions.
The bad news is that they are the same predictions that General Relativity and Quantum Mechanics make, many of which we've already tested, and is thus indistinguishable from them.
The good news is that String theory makes the same predictions as GR and QM while still being only one theory.
It is the non-compatability of GR and QM that creates the need for something like ST. If ST doesn't make a single unique prediction, but is able to explain both the quantum and relativistic worlds, then not only is that a theory, it's a great theory.
The enemies of Democracy are
Maybe after this period, people can be less childlike and some serious discussions about its strengths and weaknesses can begin.
Here's the problem though. GR and QM are both, relative to ST, extremely simple. And while ST may make the same predictions that GR and QM make, it does so in a far more complex way, without adding any extra information -- QM and GR are incompatible, but ST fails to resolve those incompatibilities in a testable way.
GR was more complex than Classical Newtonian Mechanics, but it was, essentially, a value-added theory: it explained a bunch of things that Classical Mechanics couldn't, all while remaining compatible with Classical Mechanics in places where Classical Mechanics made accurate predictions. Therefore, GR was taken to replace classical mechanics, despite the added complexity of the theory, because it was broader in scope, falsifiable, and provably more correct than the theory it replaced.
ST does not fit this mold. It is far, far, far, far more complex than either GR or QM, and makes no extra falsifiable predictions. It doesn't resolve the inconsistencies between the two. In other words, from a purely scientific perspective, it's just a hypothesis and not a particularly useful one at that.
Of course, I'm a mathematician by training and lots of interesting math has come out of ST, so for that I'm happy.
The problem is, if string theory was WRITTEN to the other theories, to describe what is happening and mathematically manipulated to come up with the same results, then it is not a theory.
If it were a theory, then string theory could independantly create a new testable hypothesis (that may be backed up by current quantum theory or relativity) and be tested based on merits of its own. Something is not a scientific theory if it merely describes what has already happened and can make no new predictions outside of that sphere. Even if the predictions were WRONG then it would still have been a theory. The world has had people creating explanations for already observed behavior for thousands of years, some have been called religions, others philosophy etc... but most didnt actually meet the criteria of being testable or (in the case of religions especially) falsifiable.
If you don't vote, you don't matter, so don't waste your time telling me your opinion
"Sterile" neutrinos, Supersymmetric particles, Kaluza-Klein particles, Energy 'leaking' into higher dimensions...
These are some of the predictions of string theories.
And they all can, to some degree, be tested empirically.
All the technology that needs to be implemented to do this isn't readily available right now, but hopefully, in coming years with experiments such as LHC and IceCube coming online, we could start to see meaningful results - Remember, it took years for empirical confirmation of General Relativity, simply due to technical limitations.
First of all it is no surprise that it resembles QM because it is QM. It assumes QM and applies it to a vibrating string, brane, etc.. But there is no new theory because there's at least a handful of different ways to do this and they're all called string theory. GR on the other hand is not as obvious. They are able to get equations that resemble Einstein's equations, but GR does NOT just pop out of it.
"Intelligent design" will never be abandoned until we're all living in the beginning of A Canticle for Liebowitz, at which point the pointy-headed mutant monks will decide that science and rationality are defeated, and the world is safe again for mad a priori assumptions that the clergy can dispense to peasants.
I read Smolen's book recently, and learned a lot of new and interesting things about string theory from it. Some problems with string theory:
Find free books.
Because it attempts to explain our reality - existence as we exist in and perceive it. Deities and the supernatural in general are by definition outside our reality, therefore not science. Even if god(s) exist, that they can mess with our reality is still an abnormality; science deals only with the natural.. uh, nature of our reality. Interference from higher powers may be fact, but it cannot be predicted (or better still, practically exploited), thus irrelevant, scienficially.
Understanding gravity, for example, allows us to navigate probes around our solar system, orbit and/or land them intact on other planets/moons/asteroids. "God did it" may be perfectly true, but it is still not a position of knowledge and understanding, it simply gets us hit in the head with fruit.
Opportunity knocks. Karma hunts you down.
Yes, and physics is full of mathematically complex theories (like schrodingers wave equations for example) but you can still predict where an artillery shell is going to land based on Newton's theories.... the theory you use is determined, in large part, by the domain you're trying to find a solution in. You wouldn't, for example, try to use QM to model the collisions of balls on a pool table. Likewise, you wouldn't use QM to try and model gravitational interactions between bodies... but the intent of string theory is that you'd be able to do either using the same framework... that alone should give some insight into the complexity of the theoretical underpinnings of it.... and explain why "breakthroughs" are taking so goddamned long. It has to maintain consistency in domains where other theories can't even represent what's going on... let alone provide the mechanism for making predictions in these domains.
Oh god, that woman is John Romero!
What I do not understand about all the negative comments on string theory is that they seem to object to it purely on the grounds that it has not yet produced a testable prediction. If there was evidence out there that it will NEVER produce a testable prediction then I would completely agree with the critics. To my knowledge this is not the case. There are certainly incredible problems to extracting a testable prediction but does that mean we should give up, pack up our bags and go home?
Sorry but sometimes physics is hard - even for physicists! Of course it might turn out in the end to be a waste of time from the physics point of view (although I'm sure even then it will leave a legacy of interesting maths) but we don't know that yet. Giving up on, from my understanding, the most promising avenue of research just because it turns out to be hard to figure it out is not good physics.
If a theory tells you what you already know to be true, then it is a retrodiction, not a prediction. There is a good reason why scientists demand that theories make predictions: A theory with a sufficient number of degrees of freedom can be made to fit any data set. For example, a polynomial of degree n can always be made to exactly fit n + 1 data points, yet may be completely unable to predict what happens between those data points.
This reminds me of Evolution, in the sense that the label "Evolution" is applied to all areas of science: Biology, Astronomy, Geology... It's as if no one would ever refute anything called "Evolution" for fear of being labeled a religious fundamentalist.
I'm uncomfortable with the automatic acceptance (at the popular level, anyway) of anything labelled evolution, without proper scientific examination. I see this all the time on documentaries, TV shows, talk shows. Some scientist will say "Oh, that's evolution." and the host will just nod his head as if it were the most obvious thing in the world.
Maybe it's just me, but if we want to keep religion out of science, we need to start with ourselves.
A year ago, nobody would force this nonsense to the table.
I can't stand popular memes! Occam started making the rounds after Jodi Foster popularized him in Contact. Ugh. The number of dumb and dumber arguments resulting from a little mis-applied knowledge was astronomical. Bubbo's Ridiculous Law, (Or whatever his name is) which states that the well-accessorized geek must close his ears upon hearing the word, "Nazi" is another.
While not quite as destructive to a healthy mental process, this cross-culture, (geek culture, that is) sudden need to lecture other geeks left and right upon the proper use of the word, "Theory", is just as annoying.
You watch. It will be mis-applied by geeks trying to knock the wind out of interesting, new ideas by declaring the ideas to be beneath even the rank of theory and therefore somehow worthy of contempt. I've seen so many people who are scared to think for themselves that unless all the ideas in their heads have been validated by somebody else, (TV or other annoying geeks with name tags), then they will shie away from them at all cost.
It's the old jr. high programming. If you are different, you will be punished through ostricization.
A cowardly geek is useless.
-FL
Yep; there are so many free parameters to superstring theory that it seems that it would be possible to create a version to suit almost any experimental observations. I know that the hope is that one day some version or other will make a useful, experimentally verifiable prediction (after all, these people are not remotely stupid; they do realise that a theory is required to make predictions), but one cannot help suspecting that, when that day comes and if the prediction turns out to be wrong, they'll just tweak one or more of the free parameters to create another one of the infinite number of possible theories so that the nre version does match the experimental result (and which will presumably make some sort of prediction for a future experiment). Wash; rinse; repeat.
I can understand why string theorists get excited about their work: there is a certain elegance to it all. But I cannot be sanguine that it will turn out to be a ToE. It may or may not be a ToN(othing); one suspects that at least some useful things will come out of it. But one cannot help thinking that it will be some much, much simpler revolution and new of looking at things that provides the real breakthrough.
Here's the problem though. GR and QM are both, relative to ST, extremely simple. And while ST may make the same predictions that GR and QM make, it does so in a far more complex way, without adding any extra information -- QM and GR are incompatible, but ST fails to resolve those incompatibilities in a testable way.
Well, I guess it depends on what you mean by "complex." Conceptually, it's actually very simple to describe, especially compared to the standard model. It's just a situation where a simple theory requires calculations too complex to carry out in order to make a definite prediction about things like cross-sections, etc. It's also not entirely true that string theory fails to be testable. It would be closer to the truth to say that it does make some generic predictions, and those predictions appear to have been falsified. It originally predicted a zero or negative cosmological constant, which turns out to be incorrect, although they've found a kludge to accomodate the observed positive value. It also predicts that supersymmetry should be a perfect symmetry of nature, which is definitely not the case experimentally, since there is no selectron with the same mass as the electron, etc.
Find free books.
Theoretically, string theory makes testable predictions and that makes the statement 'string theory makes testable predictions' a meta theory without technically fulfilling the requirements to be a theory itself and validates string theory by asserting an infinite regress which allows string theory the theory to ultimately make a testable prediction. Likewise, the statement 'theoretically, string theory makes testable predictions' can be made valid with the meta meta statement: Theoretically, in theory, string theory makes testable predictions.
Finally, all doubt can be laid to rest by validating the meta meta theory with:
In theory, theoretically, in theory, string theory makes testable predictions!
bite my glorious golden ass.
This is why you don't get it: That is behaviour which is generally considered unscientific. If you need to keep modifying your theory to explain stuff, then it's not a scientific theory. It's an ad-hoc mess of empiricism of zero real value. The rules of the game are:
1) It must be testable (falsifiable)
2) You must provide new predictions
3) You must explain previous observations, observations not used in formulating the theory., and ideally, none at all.
4) You must do so using fewer postulates (assumptions) than the previous theory.
The word you're looking for is "model". But how is this another thing? Our current understanding is a model as well. The question is whether it's a better model or not is still there and unanswered.
You have no clue. Modern chemistry is not based on any such model. It's based entirely on the standard model of physics. There is not one, not one! molecular property that can be described in anything less than a fully quantum-mechanical treatment. All of chemistry is purely due to quantum-mechanical effects.
And string theory is not the same at all, even if you'd been right. String theory is an attempt at a more basic and general theory of quantum mechanics, in the same way as classical mechanics is a limiting case of quantum theory. It is not an approximation of quantum theory, and not intended to be one.
You don't get it. The periodic table is already entirely explained from QM, and has been for some time. There's no more reason to describe it in terms of string theory than to describe the motion of billiard balls in terms of quantum mechanics: It's unnecessary because it's already explained by classical mech, and we know classical mech is a subset of quantum mech.
In the case of string theory, all they need to do is show that QM is a subset of that theory. That's not hard and it is. It forms the basic premise of their work as well as the goal. The idea is that they're going to work from part of quantum theory and relativity and somehow arrive at the whole thing. Which parts the
What a silly thing to say! String theory is scientific no matter how they arrive at their results? What if they're using a Ouija Board?
Anyway I think you misunderstood me. I don't disapprove of their methods. I am not saying you can't rely on assumptions and empirical results to form your theory. That's how all scientific theories get formed. But that methodology does automatically mean the resulting theory is scientific, or that the result is a better scientific theory.
I can go out and do an experimental observation of grass, and then formulate the theory "grass is green". This makes a prediction. It is falsifiable. But it is not a scientific theory, because it explains no more than what I'd already observed. I assume you agree to that much? You need to predict more than you assume. Part of the critisism here is based on the fear (voiced also by 't Hooft) that string theory may ultimately amount to little more than that.
It's true that you can argue it's still a scientific theory since it does explain more than it assumes, in the same way QM and relativity does. But if it makes the same number of assumptions, then it's not a better theory than those two. It's not even a new theory. It's just a useless restatement of the old one.
To give such an example: Does the Earth orbit the Sun? The heliocentric model doesn't assume more than the geocentric model. One doesn't explain more than the other. The heliocentric model is just simpler, and therefore more useful.
As I understand it, you're saying it's fine to sacrifice the goal of fewer assumptions for the goal of a more general theory in this case. That's not a view representative of what most physicists think. I'd say the goal of fewer assumptions is actually much more important in this case.
QM and relativity already explain everything we can observe so far. Likewise, what we know them not to explain (e.g. singularities in relativity) is not observable. Science is not in the business of explaining the unobservable in terms of the unknowable, and any such theory is simply unscientific, no matter how rigorous it is in terms of logic.
(Don't get me wrong, any GUT, even such an unscientific one, is still a great intellectual achievement. Just because it's not scientific knowledge doesn't mean it's not knowledge. Math is not a science, as far as I'm concerned. It's still knowledge. Logic is knowledge. Even metaphysics is knowledge - albeit not a very useful kind.)
However: It's entirely wrong to make any kind of blanket statement that string theory is unscientific. I am not doing so, nor are any critics that I know of or would consider listening to. There's no point passing judgement on a theory until there's a finished theory to judge. They're currently nowhere near that point.
But the issue of scientific rigor isn't just the aforementioned philosophical problems. There's also a more obvious social problem. String theorists are largely working in isolation from the rest of (theoretical) physics, and increasingly so. That constitutes a major warning-flag in terms of scientific rigor. Isolation leads loss of critical distance and creation of group-think. Good science is almost never done in isolation.
Because of its big goal, string theory is extremely popular and well-funded. It's a prestige subject. (and many string theorists have the big heads that go with it, another warning sign) The rest of theoretical physics is not so well funded. So a lot of people think that it's getting an inordinate amount of resources, giv
That's what you should have said, then.
It's true, you can't judge a theory until it's done. String theory is not done at all.
However, what I think they're saying, what I was saying at least, is that you can judge from the methodology used if it's going to give a useful result or not. As 't Hooft pointed out, at least some string theorists have resorted to problem-solving tactics that will end up creating more problems than they solve.
There's also a general legitimacy problem, not only within string theory (although it's particularily bad there) but within Theoretical Physics as a whole. Some areas of the field are so abstract nowadays, that few know what the heck it's all about. For instance the Bogdanoff affair (http://math.ucr.edu/home/baez/bogdanoff/). Two guys got PhDs on the basis of papers that simply didn't make any sense to anyone. And noone's quite sure whether it was sheer fraud or an honestly intended but ultimately pseudoscientific result.
But the extremely abstract nature of modern theoretical physics makes it very vunerable to this unless they interact with others. And string theory is isolated, even for being theoretical physics.
I think that even if turns out to be the right place to go, it might not be the right way to get there. Intermediate theories such as supersymmetry exist, and are not as isolated from 'real' physics.
I agree. I think most critics do as well (although perhaps not the most skeptical ones). It's more or less the only line of attack we've got towards a GUT, and we should persue it.
The real question is: How hard should we persue it? Is it getting more resources than it deserves?
It's the most important field of theory in the sense that it could provide a GUT. But it's the least important one in the sense that a GUT would have little impact on most applied physics.
Personally, I'm waiting for the Holy Grail of molecular physics: A way to solve to the molecular Schrödinger equation that scales linearly. It's been mathematically proven one exists (at least in the density-functional reformulation). We just have no clue what it is. Or even a straightforward way to find out!
Anyway.. Sorry 'bout the "no clue" remark before, that was an uncalled for. I guess I'm just a bit touchy about erroneous statements on my field of expertise.